[SOLVED] LM331 Output Frequency Increases Slowly

[imranpatel]

The output frequency of LM331 increases slowly even if the Input Voltage is constant. The output frequency increases upto 100Hz within 10 minutes and then becomes stable. Anybody having the same problem, or anybody know the solution for this undesirable behavior of LM331? Please suggest me.

 

[sequel]

hw can we help you when u have failed to show your circuit diagram...

 

[imranpatel]

Hi,

I have attached the schematic here: https://obrazki.elektroda.pl/8799701900_1351372293.png . I don't know, how to attach here.

Reference from here: **broken link removed**

In this schematic,

All resistors are Chip Resistors , +/- 1% tolerence,
C1 & C3 = Ceramic Capacitors, X7R, +/- 10%, and
C2 = Ceramic Capacitor, X7R, +/- 5%.

The response of this circuit in the beginning (at t=0) is linear, but the frequency output increase slowly within 10 minutes up to 100Hz. Also, if I continuously change the Vin voltage, the response, when first it is changed is linear, but once the voltage Vin is stable (constant), the frequency increases up to 100 Hz.

Thank you very much for your quick response.

 

[sequel]

your circuit seem rather ok. but i have had similar experience with this chip b4 but later discovered that process variation during fabrication tends to make chip with same part number and date stamp vary. i advice that u increase the value of the charge pump resistor "R3, R4" to at least 100k from which u can slowly decrease with time. also increase the value of ur load resistor "R6" to at least 4k7. as for ur C3 = 100nF; i did a lil stuff there "i used a resistor and a capacitor network there in series" i.e. i used a 10uF and a 47R resistor from PIN 1,6 to ground and it all went well....good luk as you try it out

 

[imranpatel]

Hi sequel,

Thank you very much. I will try it out.

 

[FvM]

Using X7R for the timing capacitor C2 is basically a bad idea. Must be a low T.C. NP0 ceramic or foil capacitor.

 

[imranpatel]

Hi,

Of course, what sequel said works better, as I tried it, but again the problem is resolution. Since putting 100k resistor in place of R3, and shorting R4 decreases resolution. I am already converting 50V to around 4V as Vin to LM331, so even if it changes 1 Hz frequency on this low resolution, the effect will be more. Because I have to change the frequency reading back to Voltage in Microcontroller.
One more problem is , when I built the same circuit (my original circuit), on separate board (using wires on universal board), it works very good, according to the specifications. It doesn't even change 5 Hz. But the same circuit on the prototype PCB board with other circuits doesn't seem to work, and starts behaving like I explained above. Sure I will also try with NPO timing capacitor. Can you guys help me to solve this? Why would they change the behaviour? Can it be the ground problem? I am also not putting decoupling capacitor, as my power supply has Electrolytic Capacitor of 1000 uF (at Vs pin-8 of LM331). I will try putting decoupling capacitor of 100nF at power supply, as I realized now.
Thank you very much.

 

[imranpatel]

Hi,

Finally I solved it.

I used C2 = 10nF NPO (very stable with varying temperature) capacitor in place of C2 = 10nF X7R (high temperature coefficient). As my circuit contains lot of transistors and Transformer, due to increasing temperature, the frequency was gradually increasing, which doesn't happen now after changing it to NPO.

Also, I added 47 Ohm resistor in series with C3 = 100 nF capacitor, which increased the stability further.